Cloudera Docs

Cloudera Copilot Overview

Learn how to configure and use Cloudera Copilot with Cloudera AI Inference service and Amazon Bedrock models.

Cloudera Copilot is an AI-powered coding assistant designed for seamless integration within JupyterLab ML Runtimes. With its chat interface and comprehensive code completion features, Cloudera Copilot enhances the development experience for machine learning projects. It offers compatibility with model endpoints deployed in Cloudera AI Inference service as well as Amazon Bedrock models, providing developers with flexibility and efficiency in their workflows.

Cloudera AI Inference service vs Amazon Bedrock

Cloudera AI Inference service model endpoints may be a good choice if you are concerned about proprietary data being sent to a third-party service provider. With Cloudera AI Inference service, you can run your own models and ensure that your proprietary data stays within your cloud deployment.

If you are not concerned about proprietary data being sent to a third-party service provider, and you do not expect high volumes of Cloudera Copilot usage, then using Amazon Bedrock models may be a simpler and more cost-effective solution.

Amazon Bedrock Model Support and Regional Routing

Cloudera Copilot supports a curated set of Amazon Bedrock models that are preconfigured and validated for use within JupyterLab ML Runtimes.

When configuring Copilot to use Amazon Bedrock, the available models in the Copilot UI are dynamically determined based on the following factors:
  • AWS region configuration: The specific region you have configured for Amazon Bedrock. For example, us-west-2 or eu-west-1.
  • Model availability: The specific models that are enabled and accessible within that region in your AWS account.
  • ML Runtime version: The specific version of the ML Runtime used for your active JupyterLab session.

Data Residency and Regional Routing

Starting with the ML Runtimes 2026.04.1 release, Cloudera Copilot utilizes region-specific Amazon Bedrock model identifiers, such as us.* or eu.*, rather than global.* inference profiles to ensure strict adherence to data residency and compliance requirements.

This routing mechanism ensures that all Copilot traffic remains securely within your configured AWS region. Consequently, only models that are enabled and available for your configured AWS region and account will work. Models from other regions may appear in the UI, but selecting them will result in a request failure.